His physical health is good, and I am inclined to say that he has recovered ... [but] his mind was radically changed, so decidedly that his friends and acquaintances said that he was 'no longer Gage'. (Harlow, 1868)

Observations such as these, which demonstrate that the brain and the personality are inextricable, make it clear that the object of study in psychoanalysis is somehow intrinsically connected with the object of study of neuroscience.
Freud himself recognized this fact in his neurological writings and he continued to acknowledge it throughout his psychological writings. Nevertheless, psychoanalysis developed almost completely independently of neuroscience. We all know the reason for this; despite the fact that Freud acknowledged that 'the mental apparatus ... is also known to us in the form of an anatomical preparation', as he put it (in Freud 1900, p. 536), he nevertheless always recommended that psychoanalysts should remain aloof from neuroscience. There seems at first sight to be a contradiction in this stance, but I shall clarify Freud's position in a moment.
The mere fact that psychoanalysis and neuroscience developed separately for so long points to the reality that, notwithstanding the obvious connection between the two fields, there is a great deal that separates them. Leaving aside the philosophical complexities of the mind/body relation for present purposes, it can safely be said that in practise psychoanalysis and neuroscience have separate objects of study, that they bring different methods of investigation to bear on these objects, and that the knowledge which they generate is therefore of two distinct types.
This poses obvious problems for those of us who wish to forge links between these two bodies of knowledge, as can be seen from the literature that has begun to accumulate at the interface between them over the past few decades. The first investigator to explore this relationship was, of course, Freud himself. In the mid 1890s he composed a series of drafts on the subject, one of which has survived to this day in the form of a document known to us as the 'Project for a scientific psychology'. In that work, Freud attempted to translate what was known at the time about the deeper workings of the mind, into the language of neurophysiology and anatomy. The method that Freud employed to achieve this translation was, as he himself acknowledged, one of 'imaginings, transpositions and guesses' (Freud 1954, p. 120; emphasis added). In other words, Freud relied on speculation. At that time, the gap between the knowledge that Freud had obtained by the putative method of psychoanalytical investigation about the inner workings of the mind, and the knowledge that was available to him from physiological and anatomical methods of study about the inner workings of the brain, was so large that he had no choice but to fall back on speculation when trying to bridge it. This gap played no small part in Freud's eventual abandonment of his 'Project', and his description of it as 'a kind of aberration' (ibid., p. 134). Ultimately he concluded that 'in the end I may have to learn to content myself with the clinical explanation of the neuroses' (ibid., p. 137, emphasis added; cf. Solms & Saling 1986). This remark about clinical explanation, as you will see shortly, is pregnant with implications for our science.
It is my contention that the cause of Freud's failure to integrate his clinical findings with the neuroscience of his day, was not only the paucity of neuroscientific knowledge that was available to him in the 1890s, but also the absence of a suitable method for relating the neurological and psychological data that was available. I believe further that, despite the rapid and exponential increase in our knowledge in all branches of neuroscience, every subsequent attempt to correlate psychoanalytic and neuroscientific knowledge has stumbled on the same basic problem that Freud's did, 100 years ago -- namely, the failure to develop a valid method for relating the clinical findings of psychoanalysis with the type of knowledge that is generated by the various neurological sciences. Every investigator who has written on this subject since Freud (1) -- despite the brilliance of some of their intuitions -- has relied on the same fundamental methodology that Freud did as regards the actual manner in which they correlated the two fields of inquiry, namely speculation.
These are some of the reasons why we are now in a situation, with the sudden increase of publications in this area over the past few years, where we have a number of competing and in many respects contradictory models of the neurological organization of the deep mental functions we study in psychoanalysis, without us having any rational basis for deciding between them. How are we to decide between the rival points of view? I am sure you will agree that we ought to be able to decide such questions; for, if the mind and the brain both function in regular and lawful ways, and if those functions and regularities are related to each other in similarly lawful ways -- as we have every reason to believe that they are -- then it should be possible for us to decide such questions in ordinary scientific ways.
My primary aim in this paper is to introduce you to a method by which we can achieve this ordinary scientific task. I intend to do so, firstly, by telling you something about where this method came from, secondly, how it works, and thirdly, by reporting very briefly some of the findings that this method is beginning to produce about how the deeper strata of the mind are organized neurologically.

1.

In order to do that, I want first of all to take you backwards into history, to trace the origins of psychoanalysis to a particular branch of neuroscience, and to show you how the psychoanalytic method grew out of that branch; then I want to trace subsequent developments in that field, in order to show you that it still remains the natural point of contact between our two disciplines. In the process, I hope to be able to demonstrate that -- just as we find in our clinical work -- a problem which seems insolubly complex in its present, mature form, frequently turns out to have a relatively simple structure when one traces it back to its origins.
To start at the beginning: Freud began his scientific career as a neuroanatomist, before, following a brief flirtation with psychopharmacology, he turned his attention to the problems of clinical neurology. By the time that Freud came to clinical neurology, it was still a very young discipline, which rested almost entirely on a single method. That method was known as clinico-anatomical correlation, which was carried over to the new speciality of neurology by some of the ablest practitioners of the art of internal medicine. As its name suggests, internal medicine concerned itself with the diagnosis and treatment of diseases occurring in the interior of the body, which could for that reason not be apprehended directly in the living clinical case, but rather had to be inferred from their indirect manifestations in the form of external symptoms and signs. One had to wait for the death of the patient, and the pathologist's report, before one could determine conclusively whether one's diagnosis was correct or not. But with the accumulation of experience over generations, regarding what sort of clinical presentation during life tended to correlate with what sort of pathological-anatomical findings at autopsy, it gradually became possible for internal physicians to recognize pathognomonic constellations of symptoms and signs, and thereby to predict with reasonable accuracy what the underlying disease process was, and to conduct the treatment accordingly. This was the origin of the concept of clinical syndromes, a concept with which I presume many of you are familiar.
Neurology became a separate speciality of internal medicine as it became increasingly evident, not only that the brain -- like any other organ -- was subject to its own special pathologies peculiar to its tissues, but also that damage to different parts of the brain produced a wide variety of different clinical manifestations. When Freud trained in clinical neurology in the early 1880s, this was the art that he learned; rational diagnosis and treatment of neurological diseases by the syndrome method, based on knowledge obtained by the method of clinico-anatomical correlation. In fact, we are told that Freud was a particularly gifted practitioner of this art (Jones 1953). He published a series of articles at the time attesting to his skill.
Now because, as I said at the outset, brain lesions cause mental changes, the clinico-anatomical method could be, and was, also put to another use, namely the localization of mental functions. In the early 1860s Pierre Paul Broca demonstrated conclusively that disease in a particular part of the brain -- marked by the letter A on the first diagram -- produced a highly characteristic mental symptom, namely loss of speech. On the basis of this clinico-anatomical correlation, Broca localized the faculty of speech to that small part of the brain. Ten years later, Carl Wernicke demonstrated that damage to a different part of the brain - marked by the letter B on the diagram - produced a different mental symptom, namely loss of the capacity to understand spoken language -- and he too localized that function accordingly. These two seminal discoveries were followed by a rapid series of clinico-anatomical correlations in regard to a variety of other mental functions, such as skilled movement, object recognition, and even 'intelligence'. On this basis a wide range of psychological faculties were localized within a mosaic of so-called 'centres' on the surface of the hemispheres of the brain. This was the origin of a sub-speciality within the neurological sciences, known as behavioural neurology.
We know from Freud's writings of that time that he was thoroughly versed in the methods and discoveries of this exciting new branch of science. In fact there is abundant evidence to suggest that the clinico-anatomical localization of mental functions was a subject of special interest to him (Freud 1888, 1891b, 1893-94). Clearly then, Freud was aware, before he conceived of psychoanalysis, that there was a well-established method by which it was possible to correlate mental functions on a clinical basis with the functions of particular parts of the brain. But if that was so, it raises the question, why did Freud not use this method to identify the neurological correlates of the psychological processes that he later discovered? And why don't we use it to do so today?
As I have said, Freud was an unusually gifted physician, and it didn't take him long to master the syndrome method in his diagnostic work, and the clinico-anatomical method in his ongoing research. It also did not take Freud long to discover the limits of this method. He soon came to the conclusion that it was, as he put it, just 'a silly game of permutations' (Bernfeld, 1951, p. 214). This is how it happened. Although it is true that the clinico-anatomical method was the only viable research technique available to the nineteenth-century neurologist interested in mental functions -- it was in fact used in subtley different ways within two rival schools of neurology. In the Austro-German school, within which Freud was initially trained, the emphasis fell squarely on the anatomical side of the clinico-anatomical equation. According to this school, the primary aim of neurological science was not simply to recognize which syndromes correlated with which lesions, but rather to explain the mechanism of the clinical phenomena - and thereby the corresponding normal mental functions - in anatomical and physiological terms. This approach reflected the broader ideals of the Helmholtz school of medicine.
In the rival, French school of neurology, on the other hand, the emphasis fell very much on the clinical side of the clinico-anatomical equation. According to this school, which collected around the personality of Charcot and the famous wards of the Salpêtrière Hospital, the primary task of neurological science was not so much to explain the various clinical pictures, but rather to identify, classify and describe them. The following quotation graphically illustrates the difference between these two ways of applying the clinico-anatomical method:

Charcot ... never tired of defending the rights of purely clinical work, which consists in seeing and ordering things, against the encroachments of theoretical medicine. On one occasion there was a small group of us, all students from abroad, who, brought up on German academic physiology, were trying his patience with our doubts about his clinical innovations. 'But that can't be true,' one of us objected, `it contradicts the Young-Helmholtz theory [of vision].' He did not reply 'So much the worse for the theory, clinical facts come first' or words to that effect; but he did say something which made a great impression on us: ['Theory is good; but it doesn't prevent things from existing'.] (Freud 1893a, p. 13)

This was one of Freud's favourite anecdotes.
As is well known, during his period of study at the Salpêtrière in the mid-1880s, Freud moved from being under the direct, personal influence of some of the leading figures of the Austro-German school of neurology, to being under the direct, personal influence of Charcot. This shift had a decisive influence on his thinking, and in particular, on his attitude to clinico-anatomical localization.
The reason for this shift was simple. Although the differences between the German and French schools of neurology complemented each other well in regard to most physical neurological disorders, with the one school emphasizing the anatomical and the other the clinical side of the equation, there was one group of diseases -- which were considered to fall under the domain of neurology at the time -- which threw the differences between the two approaches into sharp relief. I am referring to the neuroses, and to hysteria and neurasthenia in particular, where no demonstrable lesion of the nervous system could be found at autopsy to account for the clinical symptomatology observed during the life of the patient. This posed no serious problems for the French school: Charcot simply proceeded to describe the pathognomonic clinical syndromes of hysteria and neurasthenia, as he had done with countless other 'nervous' diseases. The neuroses were for Charcot, as Freud wrote at the time, 'just another topic in neuropathology' (ibid., p. 20). However, for the German school of neurologists, the problem was well nigh insoluble. How was one to explain in anatomical and physiological terms the mechanism of a clinical syndrome which had no pathological-anatomical basis? As a result, some German neurologists, Freud's teachers among them, developed elaborate speculative theories, while others simply declared that the neuroses were not fit subjects for serious scientific attention; if there was no anatomical lesion, there was no clinical syndrome.
During the crucial period that Freud studied under Charcot, this was the subject that most pre-occupied him. Initially Freud became a devoted pupil of Charcot, and upon his return to Vienna he expounded his views whenever and wherever he could -- much to the irritation of his old teachers. However, with increasing clinical experience, and under the influence of the English neurologist John Hughlings Jackson, Freud gradually began to edge away from Charcot and to develop a viewpoint which was at the time rather unique. Charcot was content merely to describe the clinical syndromes of hysteria and neurasthenia -- on the assumption that their pathological-anatomical correlates (which he believed had an hereditary aetiology) would eventually yield to advances in microanatomical and other laboratory techniques. Freud, on the contrary, came to the view -- some time between 1887 and 93 -- that an understanding of these clinical syndromes would never be found in pathological anatomy, or at least not by the method of clinico-anatomical correlation. He based this conclusion on two major observations, which he had first made in regard to another subject in neurology, a subject which had revealed to him the limits of the clinico-anatomical method. This subject was the problem of aphasia -- that is, precisely the subject to which the clinico-anatomical method had first been applied for the localization of mental functions -- by Broca and Wernicke -- twenty or so years before.
Freud's critical observations were the following (Freud 1891b). First, he observed that psychological faculties are complicated things, which have their own compound internal organization, and that these faculties break down according to the functional logic of their own internal construction, not according to the structural laws of cerebral anatomy. The laws of psychlogical functional systems therefore need not have any direct relationship to the structural layout of the nervous system. For this reason, Freud concluded, psychological syndromes need both to be described and to be explained in their own psychological terms. Freud's second observation, closely related to the first, was the following: psychological functions are, in their essence, dynamic processes; they arise out of a complex interplay of forces between more elementary component functions, and they are constantly re-structuring and re-adapting themselves to changing circumstances; their physiological correlates can therefore never be localized within discrete anatomical centres; they must be thought of as processes -- the dynamic resultants of interactions between the static elements of the nervous system.
It is of crucial importance for us to note that Freud first made these observations, not with reference to hysteria or any other neurosis, but rather in a study of aphasia -- that is, a syndrome which can only ever occur in the context of a definite brain lesion. In other words, these were conclusions that Freud arrived at while he was still a fully-fledged neurologist. This is underlined by the fact that he quickly went on to make similar observations in regard to non-psychological, but equally complex functions of the brain. In his writings on the disorders of voluntary movement that occur in cerebral palsy, for example, Freud went out of his way to demonstrate that they could not be localized. In a series of monographs on the subject (1891a, 1893b, 1897), like in his book on aphasia (1891b), Freud appealed to dynamic developmental factors, rather than static anatomical ones, to explain the various movement disorders in terms of specific breakdowns of the complex functional system that supports them.
It was only later that Freud applied these principles to psychopathology -- between 1893 and 1900 -- which is when psychoanalysis was born. This is a fact of critical importance, because in the next section of this paper I want to show you how these principles were subsequently developed and expanded within the field of neurology, and how a neuroscientific method for studying the cerebral organization of mental functions was eventually established on precisely these principles. That is obviously of central importance to us, in our quest for a method by which we may rejoin psychoanalysis with neuroscience.
But before I move on, let me re-cap, and summarize Freud's standpoint. He trained in the clinico-anatomical method of localizing mental functions, within the Austro-German school of neurology, which emphasized the goal of physiological and anatomical explanation. Then he shifted allegiances to the French school, which emphasized the clinical side of the equation: the elucidation of pathognomonic clinical syndromes. He used this clinical-descriptive approach to make a number of highly valued contributions to neurology, firstly in studies of aphasia, then of cerebral palsy, and lastly of neuroses. In the process of doing this work, Freud rejected the clinico-anatomical method of localizing mental functions -- indeed of localizing any complex functions -- within circumscribed anatomical centres. Freud was forced to conclude that the clinico-anatomical method could only be used to localize the most elementary functions, corresponding in the mental sphere to our primary sensory modalities (of vision, hearing, taste and so on), but that it was quite impossible to localize the neurological organization of whole mental faculties, which have supra-ordinate principles of organization based on their own internal construction, which constantly changes in the process of their development, and in their adaptation to unfolding circumstances. For Freud, such complex functions arise out of the dynamic interplay of a variety of more elementary functions. He concluded that we should conceive of this interplay as occurring between the elementary structures of the brain, and therefore forego the temptation to localize them within those elements themselves. It was obvious to Freud, schooled as he was in the art of meticulous clinical observation, that the essential factors in the aetiology and mechanism of the neuroses arose out of complex functional dynamics of this kind, and that they could therefore never be localized. This led Freud the neurologist to generalize the conclusions that he had reached in regard to speech and language and voluntary movement to the whole mental field, and to write the following fateful words in The Interpretation of Dreams, which marked the final split between psychoanalysis and the clinico-anatomical method:

I shall entirely disregard the fact that the mental apparatus with which we are here concerned is also known to us in the form of an anatomical preparation, and I shall carefully avoid the temptation to determine psychical locality in any anatomical fashion. I shall remain upon purely psychological ground, and I propose simply to follow the suggestion that we picture the instrument which carries out our mental functions as resembling a compound microscope or a photographic apparatus, or something of the kind. On that basis, psychical locality will correspond to a point inside the apparatus at which one of the preliminary stages of an image comes into being. In the microscope or telescope, as we know, these occur at ideal points, regions in which no tangible component of the apparatus is situated. (1900, p. 536, emphasis added)

What Freud retained, however, and carried over into the new field of psychoanalysis, was almost everything else he had learned as a neurologist. That is, he continued to rely on the clinical-descriptive methods of the French school of neuropathology, with its special emphasis on the careful study of the individual clinical case, and the identification of regular patterns of symptoms and signs with particular pathological significance, and he continued to explain the clinical phenomena in terms of underlying natural forces and energies, as he had been taught to do by his original masters in the Helmholtz school of medicine, and he continued to believe that these forces and energies were ultimately somehow describable in physical and chemical terms. All that he abandoned was the notion that psychological processes, which have complex and dynamic functional organizations, can be localized in discrete anatomical areas. Henceforth, rather than attempt to explain a clinical syndrome by correlating it with hypothetical damage to one or another anatomical region, as even Charcot did, Freud investigated the internal psychological structure of the syndrome and explained it by reference to a complex functional system which he assumed to be dynamically represented between the elements of the brain.
That is why Freud continued to acknowledge throughout his scientific life that the model of the mental apparatus that he devised to account for his clinical observations was a provisional construct, a system of functional relations which must be represented somehow in the tissues of the brain, and that is why he continued to insist that we in psychoanalysis should 'not mistake the scaffolding for the building', and so on. I am sure you are all familiar with Freud's many comments to the effect that psychoanalysis will someday be rejoined with neuroscience. I want only to remind you that he always insisted that this would not be possible until neuroscience developed a method which was capable of accommodating the complex, distributed and dynamic nature of the human mental process. I will read just one quotation of this sort, written in the year of Freud's death:

The psychical topography that I have developed ... has nothing to do with the anatomy of the brain, and actually only touches it at one point. [He is referring to the primary sensory modalities of the system Pcpt.] What is unsatisfactory in this picture - and I am aware of it as clearly as anyone - is due to our complete ignorance of the dynamic nature of the mental process. (Freud 1939, p. 97)

Until that was understood, Freud insisted, psychoanalysis should continue to investigate and understand the functional organization of the mental apparatus in its own terms, using a purely clinical method, disregarding its anatomical representation.
This places psychoanalysis in a very particular relationship to the neurological sciences. It places its fundamental assumptions and basic method within a well-established tradition in behavioural neurology, a tradition which has always been closely associated with the clinical-descriptive emphasis first promulgated by Charcot, and which, following Hughlings Jackson, has always rejected the notion that complex mental faculties can be concretely localized in the brain. I am referring to the dynamic school of neurology, which has been associated through the years with such outstanding physicians and theoreticians as Constantin von Monakow, Pierre Marie, Henry Head, Kurt Goldstein, Aleksandr Romanovich Luria and, most recently, Jason Brown.
The influence of this branch of neuroscience has waxed and waned over the decades. Currently it is increasing enormously, as functional imagery and computer simulation studies have revealed the fundamentally non-localizable and dynamic, 'parallel distributed processing' that underlies all mental functioning, and indeed all complex functions of the brain. The clinical emphasis of this branch of neuroscience, on the other hand, is on the wane; with the enormous strides that have been made in the use of technological auxiliary aids in medicine, the art of clinical judgement is no longer so highly valued, and the human factor in medicine is being lost. Ironically, one could say that psychoanalysis stands together with this branch of neurology as one of the last outposts of the great clinical traditions of internal medicine.
But the important point for our purposes (looking forwards rather than backwards) is that Freud carried over from neurology into psychoanalysis a basic method - namely the clinical-descriptive method, or the method of syndrome analysis as it later came to be known - and a basic conceptualization of brain-behaviour relationships - namely the anti-localizationist or dynamic conceptualization, which gives pride of place to psychological methods of analyzing mental syndromes, regardless of whether those symptoms have an organic basis or not. This method, and these basic principles, have determined the object of study of psychoanalysis, the way we go about studying it, and most important of all, the sort of knowledge that psychoanalysis generates.
Now, if we wish to integrate knowledge of this sort with knowledge about the brain, then our natural point of contact is with that branch of neuroscience which shares our basic assumptions, and out of which psychoanalysis grew; that is the dynamic school of behavioural neurology - or neuropsychology as it later came to be known. If we try to relate our clinically generated psychoanalytical knowledge with knowledge about the brain generated by fundamentally incompatible methods, or by methods which Freud explicitly rejected, then we are not only confronted by the insoluble problem of having to rely on speculation (as I said before), but we also have to recognize that we may be doing violence to the basic premises upon which our discipline was built. I am sure you will agree -- and this was always Freud's most fundamental viewpoint on the matter -- that there is little point in rejoining psychoanalysis with neuroscience, if it means that we have to abandon all that psychoanalysis stands for in the process.

2.

What I would like to do now, is describe to you one of the major developments that has occurred since Freud's death in the branch of neuroscience out of which psychoanalysis arose, because I believe that this development provides us with a method by which we can rejoin psychoanalysis with neuroscience in a way that is compatible with Freud's basic assumptions.
During the early 1920s a young Russian psychologist wrote to Freud to apply for formal recognition of a new psychoanalytical society he had formed, in the Eastern city of Kazan. This man was Aleksandr Romanovich Luria. Freud granted the recognition, and a brief correspondence ensued. A few years later, Luria moved to Moscow, and joined the Russian Psychoanalytical Society. Over a period of about ten years, Luria conducted a wide range of psychoanalytic research, published a huge number of articles, monographs, and brief reports, and conducted clinical work in a local psychiatric hospital, including (it is rumored) the analysis of Dostoevsky's granddaughter. Luria was drawn to psychoanalysis, he wrote, because it was the only branch of psychology which was both solidly rooted in natural science, and studied the living experience of real human beings.
However, the tide of political opinion soon turned against psychoanalysis in the Soviet Union, and by the early 1930s, -- fearing for his academic future, if not his life -- Luria resigned from the Russian Psychoanalytical Society, abruptly ceased all psychoanalytic activities, and delivered a penitent speech in which he admitted to his ideological mistakes, namely, according to the Party line of that time, that psychoanalysis 'biologized' human behaviour and ignored its social origins. This was a surprisingly naive remark coming from somebody with so complex an understanding of Freud's teachings, but that was not the point. Interestingly, there is evidence to demonstrate that Luria never gave up his private interest in psychoanalysis, whatever his public pronouncements. Consider for example a letter he wrote to Oliver Sacks in the mid 1970's, in which Luria described the verbal tics of a patient with Gilles de la Tourette'e syndrome as an introjection into the superego of the father's punitive voice (personal communication from Oliver Sacks to the author, March 17, 1987).
It is also very striking, in view of the charge that psychoanalysis biologized human behaviour, to observe what Luria did next (after resigning from the Psychoanalytical Society). He went to medical school, specialized in neurology, and then immediately set about studying the mental symptoms of his neurological patients. And his first piece of research in this field -- his doctoral thesis in fact -- was on the exact same subject that was preoccupying Freud when he left that field some 40 years before, that is, he studied the subject of aphasia.
When Luria finally published the results of his efforts, in 1947, in a monograph in which he scrupulously avoided the name of Freud, he proposed a theory of the cerebral representation of language which was remarkably similar to the one that Freud had proposed in 1891 (Luria 1970). I am skipping over the details here, but would like neverthtless to read you one brief quotation. Consider the striking similarity between Freud's suggestion that we view the mind as a complex optical instrument in which psychical locality corresponds to an ideal point in which no tangible component of the apparatus is situated, and the following statement by Luria: 'all attempts to postulate that ... ideas could be found in single units of the brain were as unrealistic as trying to find an image inside a mirror or behind it' (1987, p. 489). However, Luria went a step further than Freud; and this represented the essential advance that he contributed to behavioural neurology, or neuropsychology as he preferred to call it.
Luria described his approach as a 'neurodynamic' one. He used the following analogy to illustrate the principle:

Most investigators who have examined the problem of cortical localization have understood the term function to mean the 'function of a particular tissue' ... It is perfectly natural to consider that the secretion of bile is a function of the liver and the secretion of insulin is a function of the pancreas. It is equally logical to regard the perception of light as a function of the photosensitive elements of the retina and the highly specialized neurons of the visual cortex connected with them. [You will recall, this was the type of function that Freud believed could be localized.] However, this definition does not meet every use of the term function. When we speak of the 'function of respiration,' this clearly cannot be understood as the function of a particular tissue. The ultimate object of respiration is to supply oxygen to the alveoli of the lungs to diffuse it through the walls of the alveoli into the blood. The whole process is carried out, not as a simple function of a particular tissue, but rather as a complete functional system, embodying many components belonging to different levels of the secretory, motor, and nervous apparatus. Such a 'functional system' ... differs not only in the complexity of its structure but also in the mobility of its component parts. (1973, p. 27)

The same could be said of, for example, the function of digestion. Luria went on to argue that mental functions, too, can only be localized in this distributed, dynamic sense. In order to identify the different component parts which together make up the complex functional systems of the human mental apparatus, Luria devised a new method of clinico-anatomical correlation, known as 'dynamic localization'. The method works like this. If one wishes to identify the neurological organisation of a complex psychological function, one's first task is to identify all the different ways in which that function breaks down with focal neurological disease in different parts of the brain. Luria described this first step as 'qualification of the symptoms'. One starts with each of the different ways in which the function under study breaks down, and then carefully explores the psychological structure of each of these symptoms, identifying precisely in what way the functional system has collapsed in each case. This is done by using psychological methods of analysis in individual clinical cases.
The second step in Luria's method is called 'syndrome analysis'. That is, one examines what other functions are disturbed, apart from the primary function under scrutiny, in each case. Again, one relies exclusively on psychological methods of investigation, and one seeks to clarify the internal structure of these other, interconnected symptoms, in order to learn what they have in common with the function that is the primary focus of attention. In this way one identifies a single, underlying factor which can account for the full range of surface clinical manifestations.
Once you have identified the common underlying factor producing a range of psychological symptoms, you will not only have learned something about the deep psychological structure of the syndrome in question, you will also have identified the component function that is contributed by the part of the brain that is damaged in that syndrome. In other words, you will have identified the elementary psychological function of one particular part of the brain. This is a major advance.
Once one has studied, by this method, the full series of different ways in which a complex psychological faculty breaks down with damage to each part of the brain, then one will have discovered its distributed neurological organization, by identifying which parts of the brain contribute, and in what way they contribute, to the complex functional system subserving that faculty as a whole. One will not have localized that faculty in any one part of the brain, but one will have identified the various component elements between which, by dynamic functional interaction, that psychological faculty is represented.
To my mind, this method of Luria's marks a major step forward, because it enables us to identify the neurological organization of any mental function, no matter how complex, without contradicting the fundamental assumptions upon which our own discipline was built. By this method, complex psychological functions are still understood in their own, psychological terms; their dynamic nature is respected theoretically and accommodated methodologically; they are not reduced to anatomy and physiology, although their neurological distribution is laid bare; and something new is learned about their internal functional organization. By this method, a viable bridge is established between the concepts of psychology and those of anatomy and physiology and all the other branches of neurological science.
I hope I have not made the neuropsychological method of syndrome analysis sound too complicated, because it really is very simple. I truly believe that this method represents the breakthrough that Freud was waiting for. That is to say, I believe that it enables us to chart the neurological organization of everything that we in psychoanalysis know about the structures and functions of the mind.

3.

What I would like to do now, is to give you an example of how this method -- which I am claiming is the natural point of contact between psychoanalysis and neuroscience -- works in practise. I have chosen for my example a piece of research that I recently completed (Solms 1997), into the neurological organization of a mental function that is of special interest to psychoanalysis. I am referring to the function of dreaming.
Using Luria's method to study the dreams of 361 patients with neurological lesions, my research revealed that dreaming is disturbed by damage to six different parts of the brain. These regions are marked on the diagrams by the letters C, E, F, G, H and J. (Please note that the letter F refers to the same part of the brain as the letter E, but on the other hemisphere.)
Let me first of all describe the primary effects on dreaming caused by damage to each of these parts of the brain. If the brain is damaged in the regions marked by the letters C, E or F -- that is, in the mediobasal frontal region, or the inferior parietal region of either hemisphere -- the conscious experience of dreaming stops completely. This clinical fact tells us that the basic functions contributed by these three parts of the brain are fundamental to the whole process of dreaming, for when any one of them is damaged, the manifest dream is obliterated entirely. Why that should be the case is revealed an analysis of the psychological syndrome within which the loss of dreaming is embedded. I will return to that point in a moment.
First let me describe the other ways in which dreaming is disrupted by neurological disease. If the brain is damaged in the region marked by the letter G on the diagram (that is, in the ventral occipito-temporal region) then the conscious experience of dreaming persists, but the patient's dreams are devoid of any visual imagery. Strange as it may seem, patients with damage to this part of the brain have completely non-visual dreams. (Cases have also been described in which only selected aspects of visual imagery are disturbed -- such as, for example, colour imagery.)
If, on the other hand, damage is situated in the proximity of the region marked by the letter H on the diagram (that is, in the temporal-limbic region), and if the lesion is accompanied by a discharging focus (that is, by seizure activity) then the patient experiences recurring, stereotyped nightmares. These nightmares stop if the seizure disorder is brought under control.
Finally, if the damage is situated in the region marked by the letter J on the diagram (that is, in the frontal-limbic region), the patients experience a massive increase in the frequency of dreaming; they sometimes experience continual dreaming; and they have great difficulty in distinguishing between dreams and real experiences.
So, these symptoms qualify the different ways in which dreaming can be disrupted by damage to the human brain. Now, as I have said, in order to discover what the cause of the breakdown of dreaming is in each of these six instances, it is necessary to study the constellation of other psychological symptoms which accompany the changes in dreaming, following damage to each of these areas. This enables the investigator to isolate the elementary underlying factor, which is common to all of these symptoms, and which is therefore contributed by the part of the brain in question to the overall process of dreaming.
So what are the six elementary factors that are contributed by each of these parts of the brain? Unfortunately, due to limitations of time, I cannot describe the full richness of the psychological syndromes from which we have inferred these underlying factors. As a result I am going to have to over-simplify things somewhat; for expository purposes, I can only say that an analysis of the psychological syndromes associated with lesions to the six areas of the brain concerned, reveals the following basic factors. Region C contributes a general motivational factor to mental functioning. Region E contributes a factor of quasi-spatial synthesis, which is fundamental for symbolic mental operations. Region F contributes a factor of concrete spatial representation. Region G contributes a factor of revisualization, which is essential for visual mental imagery. Region H contributes an emotional arousal factor. And region J contributes a factor of selectivity, or of selective activation and inhibition, which is essential for processes such as attention, reality-testing and judgement.
So, these six factors together make up the functional system of dreaming. Or to put it differently, the process of dreaming arises out of a dynamic interaction between these six factors, which are contributed by six parts of the brain. An analysis of the special structural and functional properties of these six different brain regions, and of the dynamic relations between them, provides a basic scientific understanding of the anatomy and physiology of dreaming.
Finally, in order to arrive at a truly comprehensive account of the neurological organization of dreaming, it is also necessary to study the component functions of those regions of the brain which do not appear to be involved in the process of dreaming. This simultaneously reveals which elementary functions of the human mental apparatus are not involved in the psychological construction of dreams. For the purposes of this paper, I have marked just two of these regions on the diagram, which I believe are of particular interest.
The first of these is the core brainstem region, marked by the letter I. Although damage to this part of the brain severely disrupts the process of REM sleep, the conscious experience of dreaming persists in these patients. This suggests an unexpected dissociation between the physiological process of REM sleep and the conscious experience of dreams (Solms 1995). This dissociation is confirmed by the fact that lesions in the areas marked by the letters C, E and F on the diagram, which lead to a complete cessation of the conscious experience of dreaming, has no effect upon the physiological phenomena of REM sleep.
The other region of interest that I have drawn attention to on the diagram -- namely by the letter D -- (that is, the dorsolateral frontal convexity) is enormously important for the executive control of waking mental life and of voluntary motor activity. However, damage to this region has no effect at all on the conscious experience of dreams. This suggests, not surprisingly, that secondary process thinking and volitional motor activity have very little to do with the process of dreaming.
Now, if we take a step back and look at all of these factors together, we can arrive at a model of how the dynamic process of dreaming as a whole is organized in the tissues of the brain. On the basis of my research, I have proposed the following model, every detail of which is accessible to empirical verification, by a variety of neuroscientific methods.
It appears that dreaming is stimulated by an arousal process. The most common arousal process that stimulates a dream is the state of neurophysiological activation which regularly occurs every 90 minutes during sleep -- namely the state of REM, which is activated by the deep brainstem structures marked by the letter I on the diagram. However this is only one of the many arousal phenomena which may trigger the process of dreaming -- and it is by no means the essential one, for dreaming occurs quite normally without it. Another arousal process which can stimulate a dream, is a focal discharge in the temporal-limbic region, which is marked by the letter H on the diagram. However, this is a pathological arousal process, which cannot be bound by the dream process, and which therefore results in anxiety-dreams, or nightmares.
The next important component of the dream process is contributed by zone C. This region of the brain motivates appetitive interest in the world (that word, 'appetitive interest', is the term that modern neurobiologists use for what we would call `libidinal interest'). This region channels endogenous arousal processes in the direction of volitional motor activity. An arousing stimulus only triggers the dream work proper if it engages this quasi-libidinal brain mechanism. Parts of zone C also inhibit appetitive drives, and therefore, together with the selective structures of zone J, they deflect the arousal process away from the executive and motor systems of the brain, which are marked by the letter D on the diagram. The regions marked by the letter D are inhibited during sleep. But if the brain is damaged in zones C and J, it seems that the inhibition of these motor systems breaks down, with the result that goal-directed motor activity is instituted, and normal dreaming becomes impossible. This conceptualization of the process is supported by the fact that patients with damage to this region of the brain have severely disturbed sleep. If the brain is damaged in region J, however, the disturbance of inhibition is only partial, with the paradoxical result that there is an increase in dreaming and in dream-like thinking.
Next, assuming that there is a sufficient degree of frontal inhibition, the focus of the nocturnal arousal process shifts to the posterior systems of the brain, which regulate perceptual functions, and the higher spatial and symbolic operations which are based upon perception (zones E, F & G). This, then, becomes the primary 'scene of action' of the manifest dream. Here the three other factors that I mentioned come into play: symbolization (zone E), spatial thinking (zone F) and visual mental imagery (zone G). Among these three factors, it appears that symbolization and spatial thinking are the most important ones, for in their absence, dreaming again becomes impossible, and the whole process breaks down. Visual mental imagery is a less important factor, because the entire mental process of dreaming runs its course without it, the only difference being that the final conscious product is devoid of visual imagery. I am therefore inclined to place this factor of visual representation at the terminal end of the process of dream generation that I have described.
This overall picture suggests that dreaming is a regressive mental process, both triggered by and dependant on nocturnal states of arousal. These arousal states are channeled and inhibited by the systems controlling goal-directed behaviour. They are deflected away from the motor systems, towards the perceptual systems. The higher perceptual systems represent the arousal process in the form of symbolic and spatial syntheses, which are projected regressively onto the lower visual zones. In this way, the state of sleep is preserved. If, however, the nocturnal arousal process is excessive, such as occurs with seizures or defective frontal inhibition, then this sleep-protection mechanism fails, and the dreamer is disturbed, either by anxiety or by the innervation of volitional motor activity.
What Luria's method reveals about the neurological organisation of dreaming, therefore, is strikingly compatible with Freud's classical theory. Also, because of the centrality of dreaming in Freud's models of the mind, it provides us with a first foothold on the anatomical and physiological representation of some crucial psychoanalytic concepts, including aspects of the libido, censorship, symbolization, topographic regression, and so on. Moreover, although I don't have time to go into all these details here, by identifying the specific tissues of the brain that are involved in the different psychological components of dreaming, it becomes possible to study the finer anatomical, physiological and chemical correlates of that theory (2). That is why I am insisting that the method of dynamic localization provides psychoanalysis with a conceptual gateway to the basic neurosciences, and thereby to the enormous advances in knowledge which technological innovations in those fields have yielded in recent years. The potential benefits to psychoanalysis are so obvious that I need hardly enumerate them.
I hope that this brief and over-simplified example makes it clear enough how human mental functions are represented in the tissues of the brain, in the form of complex functional systems that arise out of dynamic interactions between a number of elementary component parts -- just like an image arises out of the compound optical instrument, as Freud suggested in The Interpretation of Dreams.
I hope also that this example shows how the method of syndrome analysis makes it possible to identify the component parts of the brain between which a complex mental function is distributed, and what the elementary contribution is that each of those parts contributes to the functional system as a whole. This is the scientific yield of the method that we have developed in neuropsychology over the past sixty years, since the death of Freud.

4.

Before I can end this paper, and rest my argument for an integration of psychoanalysis and neuroscience on the basis of this method, we must take account of the fact that the research which I have just described studied only the manifest dream process. In other words, it only studied directly the effects which damage to different parts of the brain have upon the conscious experience of dreaming, and it had to infer the underlying unconscious mechanisms from the manifest symptoms. This is because we cannot lay bare the full unconscious structure of a psychological syndrome by examining a neurological patient at the bedside, and still less by assessing him or her in a neuropsychological laboratory. In order to gain more direct access to these deeper mental strata of a patient, whether the patient has a brain lesion or not, we need to get to know the patient as a person, within an analytic relationship, in a reliable professional setting, within which we can win their confidence through tact and understanding, and by analyzing their resistances, and then unhurriedly observing the way in which the internal determinants of the symptoms gradually unfold in the transference, and by testing the hypotheses that occur to us in this regard in the form of appropriate interpretations, and observing the effects that these have upon the subsequent analytic material, and so on. In other words, we can only properly elucidate the dynamically unconscious structure of a mental symptom by means of the psychoanalytic method.
We all know that this is not the easiest way to study a psychological syndrome, but we also know that it is the only true and reliable method when it comes to those deeper aspects of mental life which neuropsychology has left unstudied, but which have always been of central concern to us in psychoanalysis, namely the dynamically unconscious structure of the human personality. In fact, the emotional resistances which conceal the internal structure of the personality probably explain why the neurological organization of this, the most important aspects by far of human mental life, have still not been systematically explored by the method of syndrome analysis. This is the scientific contribution that I believe psychoanalysis can make to neuroscience, and this is the next step that I believe we must now take.
Ironically, we owe the development of a clinical procedure for analyzing these deeper mental strata to the fact that Freud abandoned neuroscientific methods of investigation when he realized that they were (at that time) unable to accommodate the dynamic nature of human mental processes. Now it seems the time has come for us to re-introduce the fruits of his labours to the neuroscientific field out of which they originally grew. In doing so -- although I do not wish to underestimate the enormity of the task before us -- I believe that we will be able to gradually integrate psychoanalysis and neuroscience, on a solid clinical basis, in a way that is beneficial to both fields, without ignoring any of the valuable lessons which the pioneers of psychoanalysis fought so long and hard to learn.
What I am recommending, therefore, and what I believe will provide the essential cornerstone for a lasting integration of psychoanalysis and neuroscience, is a fully psychoanalytic investigation of patients with focal neurological lesions. In other words, I am recommending that we chart the neurological organization of the deepest strata of the mind, using a psychoanalytic version of syndrome analysis, by studying the deep structure of the mental changes that can be discerned in neurological patients within a psychoanalytic relationship. If I had more time I would have liked to describe the preliminary results of a study that my wife and colleague, Karen Kaplan-Solms and I began twelve years ago, using precisely these methods (Kaplan-Solms & Solms, in press). We have so far studied the subjective life of thirty-five patients with focal brain lesions, by taking them into psychoanalysis or psychoanalytic therapy. Colleagues in America, Austria, Sweden and Germany are beginning to undertake similar studies. This research is starting to reveal the neurological organization of those deeper functional systems which only the psychoanalytic method of investigation can reveal. I will report an aspect of this research in detail next term, at one of the Institute's research forums. I am afraid that tonight I only have time now to make the point the point that it is now possible, using the methods that I have described, to elucidate the neurological organization of the deepest mental functions that we have traditionally studied in psychoanalysis using purely psychopathological material.
I hope that I have conveyed this point convincingly, despite the fact that I have only been able to hint at how my way of approaching the problem actually works in practice, and that I have only been able to describe a fragment of the sort of data that it generates. Nevertheless I hope that I have at least been able to convince you of the principle that this is a worthwhile way to proceed. I will know that I really have succeeded in doing so if some of you pitch up at the research forum, and thereby demonstrate a wish to learn more. An enormous scientific effort lies before us, so I need hardly say that the more of us who involve ourselves in it, the better.

NOTES

(1) E.g., Epstein (1987, 1989, 1995), Erdelyi (1985), Frick (1982), Galin (1974), Hadley (1983, 1992), Harris (1986), Hartmann (1982), Heilbrunn (1979), Hoskins (1936), Joseph (1982, 1992), Kokkou & Leuzinger-Bohleber (1992), Levin (1991), Maclean (1962), McLaughlin (1978), Meyersburg & Post (1979), Miller (1991), Negri (1994), Olds (1992), Ostow (1954, 1955a, 1955b, 1955c, 1956, 1959), Palombo (1992), Peterfreund (1971, 1975), Reiser (1984, 1990), Schilder (1935), Schore (1994), Schwartz (1987, 1988), Stone (1977), Winson (1985), Zueler & Maas (1994).

(2) For example, closer analysis of the anatomical data reveals that the structures in zone C that are crucial for the generation of dreams are the basal forebrain fibre pathways that connect midbrain dopaminergic nuclei with mediobasal frontal cortex (the mesocortical-mesolimbic dopaminergic pathway). This suggests that whatever this fibre pathway does is critical for the function of dreaming. These are precisely the fibres that were targeted by the modified prefrontal leukotomy procedure that was so popular in the 1950s. There is evidence to suggest that modern antipsychotic medications act on this same pathway (Breggin 1980). A review of the older psychosurgical literature reveals that cessation of dreaming was a common consequence of prefrontal leukotomy (Solms 1997). Evidently, whatever it was that prevented leukotomised patients from sustaining their psychotic symptoms also prevented them from generating dreams. I am unaware of any research into the effects of modern antipsychotic medications on dreaming. However, there is considerable evidence that dopamine agonists in general (e.g. L-dopa) stimulate excessive dreaming and that dopamine antagonists (e.g. haloperidol) suppress it (Sacks 1985, 1990, 1991). If we review classical dream theory in the light of these findings, we have an empirical basis for linking the libidinal drive (or important manifestations thereof) with mesocortical-mesolimbic dopaminergic pathways. It is therefore not without interest that contemporary neuroscientists include these pathways in the 'curiosity-interest-expectancy command systems of the brain ... which instigate goal-seeking behaviors and an organisms appetitive interactions with the world' (Panksepp, 1985, p. 273).

BERNFELD, S. (1951) 'Sigmund Freud, MD. 1882-1885'. Int. J. Psycho-Anal., 32: 204.

BREGGIN, P. (1980) 'Brain disabling therapies'. In E. Valenstein (ed) The Psychosurgery Debate. San Fransisco: W.H. Freeman.

EPSTEIN, A. (1987) 'The phylogenesis of the "ego" with remarks on the frontal lobes'. Amer. J. Psychoanal., 47: 161-166.

------ (1989) 'Associations and their nature: A foundation of psychosomatic science'. J. Amer. Acad. Psychoanal., 17: 463-473.

------ (1995) Dreaming and other Involuntary Mentation: An Essay in Neuropsychiatry. Madison CT: IUP.

ERDELYI, M. (1985) Psychoanalysis: Freud's Cognitive Psychology. New York: W.H. Freeman.

FREUD, S (1888) 'Aphasia'. In M. Solms & M. Saling (eds) A Moment of Transition: Two Neuroscientific Articles by Sigmund Freud. London: Karnac, 1990.

------ & RIE, O. (1891a) Klinische Stdie über die halbseitige Cerebrallähmung der Kinder. Vienna: Moritz Perles.

----- (1891b) On Aphasia: A Critical Study. London: Imago.

----- (1893a) 'Charcot.' Standard Edition, 3: 11-23.

----- (1893b) Zur Kenntniss der cerebralen Diplegien des Kindesalters (im Anschluss an die Little'sch Krankheit. Vienna: Deuticke.

----- (1893-94) 'Amnesie' and 'Aphasie' in A. Bum & M. Schnirer (eds) Diagnostiches Lexikon für praktische Ärzte. Vienna & Leipzig.

----- (1897) Infantile Cerebral Paralysis. Miami: Miami University Press.

----- (1900) The Interpretation of Dreams. Standard Edition, 4 & 5.

----- (1939) Moses and Monotheism. Standard Edition, 23:

----- (1954) The Origins of Psycho-Analysis. New York: Basic Books.

FRICK, R. (1982). 'The ego and the vestibulocerebellar system: Some theoretical perspectives'. Psychoanal. Q., 51: 93.

GALIN, D. (1974). 'Implications for psychiatry of left and right cerebral specialization: A neurophysiological context for unconscious processes'. Arch. Gen. Psychiatr., 31: 572.

HADLEY, J. (1983). 'The representational system: A bridging concept for psychoanalysis and neurophysiology'. Int. Rev. Psycho-Anal., 10: 13.

------ (1992). 'The instincts revisited'. Psychoanalytic Inquiry, 12: 396-418.

HARLOW, J. (1848) 'Passage of an iron rod through the head'. Boston Med. Surg. J., 39: 389-393.

------ (1868) 'Recovery from the passage of an iron bar through the head'. Massachusetts Med. Soc. Publ., 2: 329-347.

HARRIS, J. (1986). Clinical Neuroscience: From Neuroanatomy to Psychodynamics. New York: Human Sciences.

HARTMANN, E. (1982). 'From the biology of dreaming to the biology of the mental'. Psychoanal. Study of the Child, 37.

HEILBRUNN, G. (1979). 'Biologic correlates of psychoanalytic concepts'. J. Am. Psychoanal. Assoc., 27: 597-626.

HOSKINS, R. (1936). 'An endocrine approach to psychoadynamics.' Psychoanal. Q., 5: 87-107.

JONES, E. (1953) Sigmund Freud: Life and Work, 1. London: Hogarth.

JOSEPH, R. (1982) 'The neuropsychology of development: hemispheric laterality, limbic language, and the origin of thought'. J. Clin. Psychol., 38: 4-33.

----- (1992). 'The limbic system: emotion, laterality and unconscious mind'. Psychoanal. Rev., 79: 405-456.

KAPLAN-SOLMS, K. & SOLMS, M. (in press) Clinical Studies in Neuro-Psychoanalysis. New York: IUP.

KOKKOU, M. & LEUZINGER-BOHLEBER, M. (1992) 'Psychoanalysis and neurophysiology: A look at case material from the two theoretical perspectives. An interdisciplinary understanding of some basic psychoanalytic concepts'. In M. Leuzinger-Bohleber, H. Schneider & R. Pfeifer (eds.) "Two Butterflies on My Head ..." Psychoanalysis in the Interdisciplinary Dialogue, pp. 133-179. Berlin, etc: Springer Verlag.

LEVIN, F. (1991). Mapping the Mind. Intersection of Psychoanalysis and Neuroscience. Hillsdale, NJ: Analytic Press.

LURIA, A. (1970) Traumatic Aphasia: Its Psychology, Syndromes and Treatment. Paris & The Hague: Mouton.

----- (1973) The Working Brain: An Intoduction to Neuropsychology. Harmondsworth: Penguin.

----- (1987) 'Mind and brain: Luria's philosophy.' In R. Gregory (ed.) The Oxford Companion to the Mind. Oxford: OUP.

MACLEAN, P. (1962). 'New findings relevant to the evolution of the psychosexual functions of the brain'. J. Nerv. Ment. Dis., 289-301.

McLAUGHLIN, J. (1978). 'Primary and secondary process in the context of cerebral hemispheric specialization'. Psychoanal. Q., 47: 237-266.

MEYERSBURG, H. & POST, R. (1979). 'An holistic developmental view of neural and psychological process: a neurobiologic-psychoanalytic integration.' Brit. J. Psychiat., 135: 139-155.

MILLER, L. (1991). Freud's Brain. Neuropsychodynamic Foundations of Psychoanalysis. New York & London: Guildford Press.

NEGRI, R. (1994) The Newborn in the Intensive Care Unit: A Neuropsychoanalytic Model. London: Karnac, The Clunie Press.

OLDS, D. (1992). 'Consciousness: A brain-centred, informational approach'. Psychoanal. Inquiry, 12: 419-444.

OSTOW, M. (1954). 'A psychoanalytic contribution to the study of brain function: 1, The frontal lobes'. Psychoanal. Q., 23:317.

----- (1955a). 'A psychoanalytic contribution to the study of brain function: 2, The temporal lobes : 3, Synthesis'. Psychoanal. Q., 24: 383.

----- (1955b). 'Behavior correlates of neural function'. Am. Scientist, 43: 127-133.

----- (1955c). 'Psychic contents and processes of the brain'. Psychosomatic Med., 17: 396-406.

------ (1956). 'Psychoanalysis and the brain'. Acta Medica Orientalia, 15: 167-176.

----- (1959). 'The biological basis of human behavior'. In A. Arieti (ed.) American Handbook of Psychiatry. New York: Basic Books.

PALOMBO, S. (1992) 'Connectivity and condensation in dreaming'. J. Amer. Psychoanal. Assoc., 40: 1139-1159.

PANKSEPP, J. (1985) 'Mood changes'. In P. Vinken, G. Bruyn & H. Klawans, Handbook of Clinical neurology, 45: 271-285.

PETERFREUND, E. (1971). Information Systems and Psychoanalysis: An Evolutionary Biological Approach to Psychoanalytic Theory. Psychol. Issues, 7 (1 & 2), Monograph 25/26.

------ (1975). 'New general theoretical frame of reference for psychoanalysis'. Psycoanal. Q., 44: 534.

REISER, M. (1984). Mind, Brain, Body: Toward a Convergence of Psychoanalysis and Neurobiology. New York: Basic.

----- (1990) Memory in Mind and Brain. What Dream Imagery Reveals. New York: Basic Books.

SACKS, O. (1985) The Man Who Mistook his Wife for a Hat. London: Duckworth.

----- (1987, personal communication) Letter to Mark Solms dated March 17, 1987.

----- (1990) Awakenings (2/ed, rev.). New York: HarperCollins.

----- (1991) 'Neurological dreams'. MD, February: 29-32.

SCHILDER, P. (1935). `Psycho-analysis of space'. Int. J. Psycho-Anal., 16: 274-295.

SCHORE, A. (1994) Affect Regulation and the Origin of the Self: The Neurobiology of Emotional Development. Hillsdale NJ: Lawrence Erlbaum Associates.

SCHWARTZ, A. (1987). 'Drives, affects, behavior - and learning: approaches to a psychobiology of emotion and to an integration of psychoanalytic and neurobiologic thought'. J. Am. Psychoanal. Assoc., 35: 467.

------ (1988). 'Reification revisited: some neurobiologically filtered views of "psychic structure" and "conflict".' J. Amer. Psychoanal. Assoc., 36 (Suppl.): 359-385.

SOLMS, M. (1995) 'New findings on the neurological organisation of dreaming: Implications for psychoanalysis'. Psychoanal. Q., 64: 43-67.

----- (1997) The Neuropsychology of Dreams: A Clinico-Anatomical Study. Mahwah: Lawrence Erlbaum Associates.

----- & SALING, M. (1986) 'On psychoanalysis and neuroscience: Freud's attitude to the localizationist tradition'. Int. J. Psycho-Anal., 67: 397

STONE, M. (1977). 'Dreams, free association, and the non-dominant hemisphere: an integration of psychoanalytical, neurophysiological and historical data'. J. Am. Acad. Psychoanal., 5: 255-284.

WINSON, J. (1985). Brain and Psyche: The Biology of the Unconscious. Garden City: Anchor & Doubleday.

ZUELER, M. & MAAS, J. (1994). 'An integrated conception of the psychology and biology of supergo development'. J. Am. Acadamy Psychoanal., 22: 195-209.